Towards Resolving Propensity Contradiction in Offline Recommender Learning

TL;DR

This paper introduces a novel adversarial learning approach for offline recommender systems that effectively addresses the propensity contradiction problem, eliminating the need for propensity estimation and improving rating prediction and ranking.

Contribution

It proposes a propensity-independent method with a theoretical error bound, overcoming limitations of existing IPS-based approaches that require MCAR data.

Findings

Outperforms existing methods in rating prediction

Achieves superior ranking metrics in experiments

Does not rely on propensity estimation or MCAR data

Abstract

We study offline recommender learning from explicit rating feedback in the presence of selection bias. A current promising solution for the bias is the inverse propensity score (IPS) estimation. However, the performance of existing propensity-based methods can suffer significantly from the propensity estimation bias. In fact, most of the previous IPS-based methods require some amount of missing-completely-at-random (MCAR) data to accurately estimate the propensity. This leads to a critical self-contradiction; IPS is ineffective without MCAR data, even though it originally aims to learn recommenders from only missing-not-at-random feedback. To resolve this propensity contradiction, we derive a propensity-independent generalization error bound and propose a novel algorithm to minimize the theoretical bound via adversarial learning. Our theory and algorithm do not require a propensity estimation procedure, thereby leading to a well-performing rating predictor without the true propensity information. Extensive experiments demonstrate that the proposed approach is superior to a range of existing methods both in rating prediction and ranking metrics in practical settings without MCAR data.